Deep Convolutional Neural Networks for Massive MIMO Fingerprint-Based Positioning

Joao Vieira; Erik Leitinger; Muris Sarajlic; Xuhong Li; Fredrik Tufvesson

doi:10.1109/PIMRC.2017.8292280

Deep Convolutional Neural Networks for Massive MIMO Fingerprint-Based Positioning

Joao Vieira, Erik Leitinger, Muris Sarajlic, Xuhong Li, Fredrik Tufvesson

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

Abstract

This paper provides an initial investigation on the application of convolutional neural networks (CNNs) for fingerprint-based positioning using measured massive MIMO channels. When represented in appropriate domains, measured massive MIMO channels have a sparse structure which can be efficiently learned by CNNs for positioning purposes. We evaluate the positioning accuracy of state-of-the-art CNNs with channel fingerprints generated from a channel model with a rich clustered structure: the COST 2100 channel model. We find that moderately deep CNNs can achieve fractional-wavelength positioning accuracies, provided that an enough representative data set is available for training.

Original language	English
Title of host publication	28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC 2017.
Publisher	IEEE - Institute of Electrical and Electronics Engineers Inc.
DOIs	https://doi.org/10.1109/PIMRC.2017.8292280
Publication status	Published - 2018 Feb 15
Event	28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 2017 - Hotel Bonaventure, Montreal, Canada Duration: 2017 Oct 8 → 2017 Oct 13

Conference

Conference	28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 2017
Abbreviated title	PIMRC
Country/Territory	Canada
City	Montreal
Period	2017/10/08 → 2017/10/13

Subject classification (UKÄ)

Communication Systems

Access to Document

10.1109/PIMRC.2017.8292280

Cite this

Vieira, J., Leitinger, E., Sarajlic, M., Li, X., & Tufvesson, F. (2018). Deep Convolutional Neural Networks for Massive MIMO Fingerprint-Based Positioning. In 28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC 2017. IEEE - Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/PIMRC.2017.8292280

@inproceedings{4e3ace374634414ebe93a26fd6d96e5d,

title = "Deep Convolutional Neural Networks for Massive MIMO Fingerprint-Based Positioning",

abstract = "This paper provides an initial investigation on the application of convolutional neural networks (CNNs) for fingerprint-based positioning using measured massive MIMO channels. When represented in appropriate domains, measured massive MIMO channels have a sparse structure which can be efficiently learned by CNNs for positioning purposes. We evaluate the positioning accuracy of state-of-the-art CNNs with channel fingerprints generated from a channel model with a rich clustered structure: the COST 2100 channel model. We find that moderately deep CNNs can achieve fractional-wavelength positioning accuracies, provided that an enough representative data set is available for training.",

author = "Joao Vieira and Erik Leitinger and Muris Sarajlic and Xuhong Li and Fredrik Tufvesson",

year = "2018",

month = feb,

day = "15",

doi = "10.1109/PIMRC.2017.8292280",

language = "English",

booktitle = "28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC 2017.",

publisher = "IEEE - Institute of Electrical and Electronics Engineers Inc.",

address = "United States",

note = "28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 2017, PIMRC ; Conference date: 08-10-2017 Through 13-10-2017",

}

Vieira, J, Leitinger, E, Sarajlic, M, Li, X & Tufvesson, F 2018, Deep Convolutional Neural Networks for Massive MIMO Fingerprint-Based Positioning. in 28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC 2017.. IEEE - Institute of Electrical and Electronics Engineers Inc., 28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 2017, Montreal, Canada, 2017/10/08. https://doi.org/10.1109/PIMRC.2017.8292280

Deep Convolutional Neural Networks for Massive MIMO Fingerprint-Based Positioning. / Vieira, Joao; Leitinger, Erik; Sarajlic, Muris et al.
28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC 2017.. IEEE - Institute of Electrical and Electronics Engineers Inc., 2018.

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

TY - GEN

T1 - Deep Convolutional Neural Networks for Massive MIMO Fingerprint-Based Positioning

AU - Vieira, Joao

AU - Leitinger, Erik

AU - Sarajlic, Muris

AU - Li, Xuhong

AU - Tufvesson, Fredrik

PY - 2018/2/15

Y1 - 2018/2/15

N2 - This paper provides an initial investigation on the application of convolutional neural networks (CNNs) for fingerprint-based positioning using measured massive MIMO channels. When represented in appropriate domains, measured massive MIMO channels have a sparse structure which can be efficiently learned by CNNs for positioning purposes. We evaluate the positioning accuracy of state-of-the-art CNNs with channel fingerprints generated from a channel model with a rich clustered structure: the COST 2100 channel model. We find that moderately deep CNNs can achieve fractional-wavelength positioning accuracies, provided that an enough representative data set is available for training.

AB - This paper provides an initial investigation on the application of convolutional neural networks (CNNs) for fingerprint-based positioning using measured massive MIMO channels. When represented in appropriate domains, measured massive MIMO channels have a sparse structure which can be efficiently learned by CNNs for positioning purposes. We evaluate the positioning accuracy of state-of-the-art CNNs with channel fingerprints generated from a channel model with a rich clustered structure: the COST 2100 channel model. We find that moderately deep CNNs can achieve fractional-wavelength positioning accuracies, provided that an enough representative data set is available for training.

UR - https://arxiv.org/abs/1708.06235

U2 - 10.1109/PIMRC.2017.8292280

DO - 10.1109/PIMRC.2017.8292280

M3 - Paper in conference proceeding

BT - 28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC 2017.

PB - IEEE - Institute of Electrical and Electronics Engineers Inc.

T2 - 28th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 2017

Y2 - 8 October 2017 through 13 October 2017

ER -

Deep Convolutional Neural Networks for Massive MIMO Fingerprint-Based Positioning

Abstract

Conference

Subject classification (UKÄ)

Access to Document

Other files and links

Fingerprint

Algorithms and Proofs of Concept for Massive MIMO Systems

Cite this